The post seems to spend quite a lot of time on the V language and discussing it, so I got curious and put in some of the examples into V for windows. The results seem to be different, but my knowledge on V is too low to understand the implications fully.
The "No null" experiment does not give a segmentation faul, instead it prints &nil. That seems to be effectively a null value to me though. Adding another layer of null dereferencing creates an access violation/sefault.
The undefined behaviour for integer overflow is the exact same as in the post. (edit: and in the github issues it is marked as "won't fix" it seems V does not care about creating undefined behavior in the generated C code, just that the behavior is defined within V.)
The overwriting of len for the array seems to be fixed as the capacity is increased to the value of len seems to return nil references now.
The immutable->mutable trick seems to be caputred by the type checker.
The remarks on included libraries seems to hold, as well as the "simple webserver" being 2MB when compiling (although openssl seems to be included now).
I have no motivation to redo all the other experiments in the blog post. And your response as well as the small checks leave me not-very-impressed with V, as I can't take any of the claims on the website seriously like that. If you claim it is a v0.1 or v0.3 language, sure, make a list of features that actually exist or a roadmap when they are reached. This is especially true for the "novel memory management" which seems optimistically to assume it worked before when a changenote is included that the compiler doesn't leak memory anymore by default. That is not particularly instilling confidence, although it's the right direction.